The "Reseed Identity Columns" option has stopped working

MEGADETH20650

Had I known that this was not a fully tested release, I would not have downloaded it. But when I started SQL Data Compare yesterday, it told me that an upgrade was available, so I dutifully downloaded it.

Is there a way for me got get back to the previous version?

David Atkinson

The 'reseed' identity has changed its behavior due to numerous complaints we had about it causing failed inserts on the target. The current behavior is to reseed without any arguments, rather than explicitly, which lets SQL Server set the appropriate seed value, rather than copying it naively from the source database.

If you could describe exactly what your use case is, allowing us to reproduce the scenario in-house, we would certainly consider restoring it as a further option.

Kind regards,
David Atkinson
Product Manager
Red Gate Software

MEGADETH20650

I have 2 databases with identical schema. One contains only lookup data - this is what we would deploy to a new customer. The other database is the operational database - the one that we test our software against.

To prevent the operational database from growing too large, or when trying to isolate a bug in the data services, I sync its data with the the database that contains the lookup data. This results in non-lookup data being deleted.

Prior to last week's update, SQL Data Compare would reset the Identity columns to start at 1. Ever since the update, the Identity columns are not reset, and subsequent Inserts will not start at 1. Unfortunately this causes a number of our unit tests to fail.

In the project options, the "Reseed Identity Columns" option still exists, but it doesn't do anything.